BRugs (version 0.9-2.1)

BRugsFit: BRugs' meta function

Description

This function takes model, data and starting values as input and automatically runs a simulation in BRugs.

Usage

BRugsFit(modelFile, data, inits, numChains = 3, parametersToSave,
    nBurnin = 1000, nIter = 1000, nThin = 1, coda = FALSE,
    DIC = TRUE, working.directory = NULL, digits = 5, seed=NULL,
    BRugsVerbose = getOption("BRugsVerbose"))

Value

If coda is set to TRUE, an mcmc.list object as known from the coda package is returned, otherwise a list containg components

Stats

A data frame containing sample statistics. See samplesStats.

DIC

The DIC statistics, if DIC=TRUE, else NULL. See dicStats.

Arguments

modelFile

File containing the model written in OpenBUGS code, an R function that contains a BUGS model that is written to a temporary model file (see tempfile) using writeModel.

data

Either a named list (names corresponding to variable names in the modelFile) of the data for the OpenBUGS model, or a vector or list of the names of the data objects used by the model. In these cases data are written into a file data.txt into the temporary directory of the current R session.

If a filename of an existing file is given, data are read from that file.

inits

A list with numChains elements; each element of the list is itself a list of starting values for the OpenBUGS model, or a function creating (possibly random) initial values. In these cases inits are written into files inits1.txt, ..., initsN.txt into the temporary directory of the current R session.

If a vector of filenames of existing files is given, inits are read from those files. Alternatively, if inits is not specified, initial values are generated by OpenBUGS.

numChains

Number of Markov chains (default: 3).

parametersToSave

Character vector of the names of the parameters to save which should be monitored.

nBurnin

Length of burn in (before nIter iterations start).

nIter

Number of iterations (without burn in).

nThin

Every nThin-th iteration of each chain is stored.

coda

Determines the output format: if FALSE (default), a list containing sample and DIC statistics is returned. If TRUE, an mcmc.list object as known from the coda package is returned.

DIC

Logical, whether to calculate and return the DIC.

working.directory

Sets working directory during execution of this function; data, inits and other files are written to / read from this directory if no other directory is explicitly given in those arguments. If NULL, the current working directory is chosen.

digits

Number of significant digits used for OpenBUGS input, see formatC.

seed

Integer value from 1 to 14 defining the state of the random number generator - default is to not specify the state (see modelSetRN).

BRugsVerbose

Logical, whether BRugs is supposed to be verbose. This can be controlled for the whole BRugs package by by the option ‘BRugsVerbose’ (see options) which is set to TRUE by default.

See Also

BRugs, help.WinBUGS. Andrew Gelman proposes some print and plot methods that can be accessed by the openbugs (and bugs) and as.bugs.array functions in the CRAN package R2WinBUGS.

Examples

# \donttest{
## Prepare the example files in a temporary directory
exfiles <- dir(options()$OpenBUGSExamples, pattern="^Rats.*txt$", full.names=TRUE)
ok <- file.copy(exfiles, tempdir())
BRugsFit(data = "Ratsdata.txt", inits = "Ratsinits.txt",
    para = c("alpha", "beta"), modelFile = "Ratsmodel.txt",
    numChains = 1,
    working.directory = tempdir())
# }